[
https://issues.apache.org/jira/browse/NIFI-5619?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jeff Storck updated NIFI-5619:
------------------------------
Summary: Node can join the cluster with a processor running when cluster
state for that processor is stopped (was: Node can join the cluster with
processors running when cluster state for that processor is stopped)
> Node can join the cluster with a processor running when cluster state for
> that processor is stopped
> ---------------------------------------------------------------------------------------------------
>
> Key: NIFI-5619
> URL: https://issues.apache.org/jira/browse/NIFI-5619
> Project: Apache NiFi
> Issue Type: Bug
> Affects Versions: 1.8.0
> Reporter: Jeff Storck
> Priority: Major
> Attachments: Screen Shot 2018-09-20 at 4.53.44 PM.png, Screen Shot
> 2018-09-20 at 4.54.07 PM.png, Screen Shot 2018-09-20 at 4.54.15 PM.png
>
>
> When the cluster state for a processor is stopped, it is possible for a node
> to join the cluster with that particular processor running. The UI will show
> the processor as stopped, but going to the cluster summary and viewing the
> node-specific status of that processor will show it's running on the node
> that joined.
> To reproduce this:
> # Build the Node Decommissioning [PR
> 3010|https://github.com/apache/nifi/pull/3010]
> # Create a two-node cluster, does not need to be secure
> # Create a flow with a GenerateFlowFile processor connected to an
> UpdateAttribute processor
> # Start the GenerateFlowFile processor to let flowfiles queue in the
> connection to UpdateAttribute
> # Disconnect the second node from the cluster
> # Decommission the second node from the cluster
> # Restart the node that was decommissioned from the cluster
> # Once the node is reconnected, in the UI the GenerateFlowFile processor will
> look like it is stopped
> # Go to the Cluster Summary UI and click on the View Processor Details icon
> which will show the processor still running on the node that was reconnected
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
